Learning graphs to model visual objects across different depictive styles

Q. Wu, H. Cai, P. Hall

Research output: Chapter or section in a book/report/conference proceedingChapter or section

16 Citations (SciVal)
314 Downloads (Pure)


Visual object classification and detection are major problems in contemporary computer vision. State-of-art algorithms allow thousands of visual objects to be learned and recognized, under a wide range of variations including lighting changes, occlusion, point of view and different object instances. Only a small fraction of the literature addresses the problem of variation in depictive styles (photographs, drawings, paintings etc.). This is a challenging gap but the ability to process images of all depictive styles and not just photographs has potential value across many applications. In this paper we model visual classes using a graph with multiple labels on each node; weights on arcs and nodes indicate relative importance (salience) to the object description. Visual class models can be learned from examples from a database that contains photographs, drawings, paintings etc. Experiments show that our representation is able to improve upon Deformable Part Models for detection and Bag of Words models for classification.
Original languageEnglish
Title of host publicationComputer Vision – ECCV 2014
Subtitle of host publication13th European Conference, Zurich, Switzerland, September 6-12, 2014, Proceedings, Part VII
EditorsDavid Fleet , Tomas Pajdla , Bernt Schiele , Tinne Tuytelaars
Place of PublicationCham, Switzerland
Number of pages16
ISBN (Print)9783319105833
Publication statusPublished - 22 Sept 2014
Event13th European Conference on Computer Vision, ECCV 2014; Zurich - Zurich , Switzerland
Duration: 6 Sept 201412 Sept 2014

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)


Conference13th European Conference on Computer Vision, ECCV 2014; Zurich


Dive into the research topics of 'Learning graphs to model visual objects across different depictive styles'. Together they form a unique fingerprint.

Cite this